ASPECTOS TURÍSTICOS

In all of the six species examined in this study, reproductive behaviour was observed over at least a 5 month period from mid-August to mid-January. Although a more detailed histological study is needed, this period probably coincides with the duration of the spawning season. Such prolonged spawning periods appear to be common in

temperate reef fish, including labrids, e.g. Notolabrus celidotus (Jones 1980), and Centrolabrus exoletus, Crenilabrus melops, Ctenolabrus rupestris, Labrus berygylta, and Labrus mixtus (Costello 1991), and monacanthids, e.g. Monacanthus hispidus (Hildebrand and Cable 1930), and have been reported for many New Zealand reef fishes by Doak (1972) and Thompson (1981), and some Tasmanian species by Gunn and Thresher (1991). The period during which spawning was observed in this study

coincides with rising water temperatures and high levels of primary production, factors that could enhance the survival and growth of larval fishes. By spawning over such a long period these species can also ensure that at least at some time during the

spring/summer period some of their offspring will be present in the water column during a peak in primary or secondary production. The advantages of such a spawning

strategy were highlighted in a study by Thresher et al. (1989) who found that transient pulses in primary production in Tasmanian waters usually preceded peaks in

recruitment success of Heteroclinus sp. They suggested that such peaks in primary production, and associated peaks in secondary production, can have a profound affect on population ecology, especially recruitment success, as it may only be at such times that the availability of food in the water column rises above the threshold needed for maintenance and growth of newborn fish larvae. Species with a prolonged spawning period are much less likely to suffer the same extent of interrannual variability in recruitment success as those that only spawn once and over a short period. Comparison of differences in the variability in recruitment success of synchronous and sequential spawning species with short and long spawning periods may prove to be a fruitful area for further research.

No evidence of demersal spawning was obtained for any of the species in this study, even though such behaviour is common in northern temperate labrids (Costello 1991) and is thought to be common amongst the monacanthids (Barlow 1987, Nakazono and Kawase 1993). The absence of demersal spawning in the labrids in this study, as well as those in New Zealand waters (Thompson 1981) suggests that the European benthic spawning species are unusual. There is no evidence of a general trend towards demersal spawning in temperate labrids from the pelagic spawning found in most tropical

species. Furthermore, demersal spawning would not normally be predicted for such large fish under the size disadvantage hypothesis of Thresher (1984). This hypothesis suggests that while small fish may tend eggs in easily defended locations such as shells and small cracks in the reef, this becomes increasingly more difficult to do with

increasing size due to decreasing agility and manoeuvrability. This hypothesis may also explain why the two large monacanthids in this study were not demersal spawners, although there are still too few studies of species in this family to ascertain whether pelagic spawning is, in fact, unusual.

3.4.2 Labrid sexual systems

Three of the four labrids examined in this study were identified as protogynous hermaphrodites, while Notolabrus fucicola is a secondary gonochorist. No primary or pre-maturational IP males were found in any of the hermaphrodite species, and where IP males were found, they were in the size range where sexual inversion would be

expected, with the exception of three N. tetricus individuals collected at Ninepin Point in September 1990 (Fig. 3.2). It is likely that these particular IP males resulted from premature sex inversion, perhaps due to a temporary absence of larger males and females. At the time the collection was made, the site was being heavily netted by amateur fishermen prior to its declaration as a marine reserve, with the mesh sizes used selectively removing the larger fish.

Robertson and Choat (1974) proposed a model that predicted the presence of IP primary males in species with "loose" sexual systems which, when applied to labrids, would include species where harems are not maintained and males do not have the opportunity to become familiar with, and control, all the females in their territory. Under this model, all the hermaphrodite species in this study would be expected to have IP males, as all had a "loose" system, where the home ranges of females overlapped the territories of several males, thus limiting the control that males may exert. As no IP males were found, it appears that these three species, as well as Labrus bergylta (Dipper and Pullin 1979) are exceptions to the model. Another partial exception to this model is

Notolabrus celidotus (Jones 1980) where primary males do not occur but a small proportion of IP pre-maturational males are present. Jones suggested that this may be a successful alternative strategy to primary males as it provides more phenotypic

plasticity and a potentially greater response to the environmental conditions which may determine the fitness of a particular sex.

A similar phenotypic plasticity may explain the absence of primary males in Pictilabrus laticlavius, Pseudolabrus psittaculus and Notolabrus tetricus. If sex inversion in these species is determined primarily by social factors such as the density of males and an individuals position in a size structured hierarchy, then such a system may provide a more than sufficient response to changing environmental conditions, negating the need for IP males. This would be particularly the case if sex change can be rapid when necessary. Certainly this type of sexual transition process is common in the labrids, and was first documented by Robertson (1972) in Labroides dimidiatus, where transition of the most dominant female within a harem, from female to fully functional male, takes only 14 days from when the dominant male is removed.

In species with such a flexible system, living within an environment where males are able to maintain stable territories, initial phase males are unlikely to make a useful contribution to the reproductive success of a population, and would not be expected to form a significant proportion of the population. However, in species where sex change is determined by size or age, selection may favour the retention of some IP males as their fitness would increase dramatically during times of high mortality of older and larger fish. This may be particularly the case for species living in temperate

environments which tend to be less stable than tropical ones and where occasional storms may have a great impact on population structure.

Further work is needed to determine the factors controlling sex inversion in both monandric and diandric species of labrids before more concrete models regarding the presence of IP males can be proposed. This work should particularly focus on

monandric temperate species such as Labrus bergylta, Pictilabrus psittaculus,

Pseudolabrus psittaculus and Notolabrus tetricus, and diandric temperate species with IP primary males such as Labrus mixtus and IP secondary males such as Notolabrus celidotus, so that any new model predicting the abundance of IP males can incorporate

the effects of social system "looseness" as well as latitude, and the degree of social control of sex inversion.

Notolabrus fucicola was unusual in that it is a secondary gonochorist, having a sexual strategy presently described for only one other species of labrid, Oxyjulis californica, by Diener (1976). This sexual system is obviously derived from protogynous ancestors, and must be the result of strong selection against TP males in favour of IP males. Some clues as to how this may have evolved in N. fucicola may be given by its preferred habitat. This species inhabits shallow waters along high energy coastlines. Because of the unpredictable nature of such a habitat, the territorial behaviour typical of this genus, and labrids in general, would be difficult to maintain, particularly during storms when fish need to move to deeper water to avoid being killed. The lack of territorial behaviour in this species is documented in Chapter 2. Without the social control that territorial behaviour enables, the proportion of IP males would be expected to increase, as predicted by Robertson and Choat (1974), or the average size of TP males would decrease due to lack of inhibition of sex change by dominant males. The resulting reduction in both the abundance and size of females would be maladaptive, and it would be expected that selection would favour a convergence with a gonochoristic system, where sexes were genetically determined. Such a system appears to be an optimal strategy for non-territorial reef fishes, as it is the system used by the majority of them.

In a book on the coastal fish of north-eastern New Zealand, Thompson (1981) reported that in New Zealand a small proportion of mature N. fucicola females change sex. If the observations reported by Thompson (1981) are correct, the mode of reproduction in the New Zealand population differs from that observed in theTasmanian population. Such a difference between populations could be explained by their physical isolation by

distance (limiting gene flow), with the New Zealand population representing an earlier and intermediate stage in the evolution of this species from an hermaphroditic ancestor. It may also explain the confusion in the literature on the sexual system of this species.

4.4.3 Monacanthid sexual system:

Both Meuschenia australis and Penicipelta vittiger are gonochoristic. This was not unexpected as hermaphroditism has not previously been reported in this family or the related balistids (Thresher 1984). The initial confusion as to whether or not these

species were gonochoristic arose because of the highly variable coloration found in both species, and the fact that differences in secondary sexual characteristics such as

coloration are not readily evident in P. vittiger until sexual maturity is approached. While the sexes of 1+ juveniles of this species could be recognised during the breeding season due to the intensification of slight differences in external coloration, this was not the case in March/April once these differences faded. Mature fish of both species however, were distinctly sexually dichromatic, and, in the case of P. vittiger, sexually dimorphic. A less distinct sexual dimorphism was also observed in mature M. australis, with females being deeper in the body with respect to length, although this remains to be described in a taxonomic study.

There is a significant difference between the sexes in the growth rates of both species (Chapter 3). Males have a greater length at age by 2+, and generally attaining greater maximum length overall, although in P. vittiger growth rate differences were not

evident in 1+ juveniles in February, and probably develop as maturity is approached and females apportion a greater proportion of their resources to reproductive effort. These differences between the sexes appear to be a characteristic feature of the monacanthids (Thresher 1984, Nakazono and Kawasee 1993) although in some species these

Chapter 5